摘要
首次提出了基于多岛遗传算法进行生物组织三维温度场无损重构的重要思想,把复杂的生物热传导反问题的求解转换为正问题的求解过程,并通过实验对该重构思想的可行性和可靠性进行了验证。以生物组织内点热源的位置P(x,y,z)和温度t为优化变量,把同一表面各个对应点的实验温度值和计算温度值相减并取绝对值之和,以此为目标函数逐次迭代。目标值越小,则当前变量,即热源位置和温度值最优。多岛遗传算法可以很好地应用于生物组织三维温度场的重构,无需提取生物组织全部的表面温度数据,为热传导反问题的研究提供了一条具有借鉴性的方法与思路。
The nondestructive reconstruction of three-dimensional（3D）temperature field in biological tissue is always an important problem to be resolved in biomedical engineering field.This paper presents a novel method of nondestructive reconstruction of 3Dtemperature field in biological tissue based on multi-island genetic algorithm（MIGA）.By this method,the resolving of inverse problem of bio-heat transfer is transformed to be a solving process of direct problem.An experiment and its corresponding simulation were carried out to verify the feasibility and reliability.In the experiment a high purity polypropylene material,whose thermophysical parameters were similar to the fat tissue being tested,were adopted so that it could avoid the negative results created by the other factors.We set the position P（x,y,z）as the point heat source in the biological tissue and its temperature t as optimization variable,got the experimental temperature values of the points in a module surface,subtracted them from the corresponding simulating temperature values in the same module surface,and then took the sum of absolute value.We took it as the objective function of successive iteration.It was found that the less the target value was,the more optimal the current variables,i.e.the heat source position and the temperature values,were.To improve the optimization efficiency,a novel establishment method of objective function was also provided.The simulating position and experimental position of heat source were very approximate to each other.When the optimum values are determined,the corresponding 3D temperature field is also confirmed,and the temperature distribution of arbitrary section can be acquired.The MIGA can be well applied in the reconstruction of 3Dtemperature field in biological tissue.Because of the differences between the MIGA and the traditional numerical methods,we do not have to acquire all the data of surface.It is convenient and fast,and shows a prosperous application future.
出处
《生物医学工程学杂志》
EI
CAS
CSCD
北大核心
2016年第4期666-673,共8页
Journal of Biomedical Engineering
基金
国家重大科学仪器设备开发专项资助项目(2012YQ160203)
湖北科技学院博士启动基金资助项目(BK1528)
关键词
多岛遗传算法
三维温度场
目标函数
集成优化
multi-island genetic algorithm
three-dimensional temperature field
target function
integrated optimization
